When, If Ever, Is Voting Speech?

By Mike Dorf


My post yesterday mentioned but did not get into the merits of the dispute in Nevada Comm'n on Ethics v. Carrigan between, on the one hand, Justice Alito, and, on the other hand, the rest of the Court, speaking through Justice Scalia. They divided over whether the casting of a legislative vote is an exercise of the "freedom of speech" for purposes of the First Amendment.  Alito said it is; Scalia and the rest of the Court said it is not.  Who's right?

Before answering that question, it's important to clarify why the question matters. The Supreme Court's cases sometimes appear to say that whether a person intended to speak is not relevant; what matters is whether the government, penalizes people because of the message they communicate.  Thus, a law that targets the communicative element of flag burning violates the First Amendment (as the Court held in Texas v. Johnson), whereas a law that forbids people from destroying their draft cards for (what the Court somewhat disingenuously accepted as) administrative purposes does not violate the First Amendment, even if the particular draft card burner intends to express a message by burning the draft card (as the Court held in United States v. O'Brien).  Indeed, subject to an exception I'll describe below, the Court has even said that wholly unprotected speech--such as fighting words--cannot constitutionally be targeted by the government if the targeting is based on hostility to the message conveyed by the unprotected speech (as it supposedly was in R.A.V. v. St. Paul).


In light of the foregoing, one might think that it makes no difference whether casting a legislative vote is speech. The key question, one might think, is whether, in forbidding the casting of particular legislative votes, the government targets the forbidden votes based on the messages they express.  However, in Nevada Comm'n, Justice Scalia rejects that view.  He says that the R.A.V. principle--protecting even wholly unprotected speech under some circumstances--only comes into play where the government targeting is viewpoint-based.


As an initial matter, I should say that this strikes me as a narrowing of R.A.V. (which was also written by Justice Scalia, albeit nearly two decades ago). There, the Court said that the particular ordinance was viewpoint-based, but the thrust of the R.A.V. opinion was that any content-based restrictions on speech, including proscribable speech like fighting words, must be subject to strict scrutiny under the First Amendment--except where the basis for the content restriction is the same as the grounds for making the category of unprotected speech proscribable in the first place. To read R.A.V. as only applying to viewpoint-based restrictions--as the Nevada Comm'n majority opinion arguably does--is to narrow R.A.V. considerably.

Be that as it may, Justice Alito does not disagree with Justice Scalia and the balance of the Court on this point. Alito is willing to accept that if legislative voting is not speech then, like proscribable speech, it only receives protection against viewpoint-based restrictions.  Or at least Justice Alito doesn't challenge the majority on its reading of R.A.V. Instead, his challenge concerns legislative voting itself.

According to Justice Scalia, legislative voting is "nonsymbolic conduct engaged in for an independent governmental purpose"--namely, legislation--even though individual legislators may, in particular circumstances, wish to communicate a message with their votes or with their votes in combination with statements. By contrast, Justice Alito says: "Voting has an expressive component in and of itself." So who's right?

I must say that I find the disagreement somewhat bewildering without criteria for deciding when some category of acts is inherently communicative. Virtually any act can be communicative or non-communicative, depending on the surrounding circumstances, the intentions of the actor/speaker, and the understanding of the audience. One could recite the words of the Constitution for a wholly non-communicative purpose--to cause an avalanche, say.  One could drive a car as a form of expression--in a slow-moving parade, say. And one could engage in non-communicative conduct--such as opening an umbrella--with the effect of unintentionally communicating information to others: here, the fact that it is raining.

What might Justices Scalia and Alito be disagreeing about when they disagree about whether legislative voting is inherently communicative? One possibility is a kind of statistical average: One could turn on the lights as a form of communication but the vast majority of light-turning-on is non-expressive, so we say the act is inherently non-expressive; one could sing a song in a foreign language without knowing the meaning of the words, simply as a means of staying awake, but the vast majority of instances of singing are expressive, so we would say music is inherently expressive. Yet neither Justice Scalia nor Justice Alito seems to have a statistical view in mind. Justice Scalia seems to think that legislative voting is inherently non-communicative because the purpose for which it was established--the enactment of legislation--is inherently non-communicative.  That's an unsatisfactory or at least unsatisfactorily explained view.  Among other things, it doesn't grapple with the fact that much legislation is itself symbolic.  But at least we can get an inkling of why Justice Scalia thinks legislative voting is inherently uncommunicative.  Justice Alito doesn't even suggest why he thinks that legislative voting is inherently communicative; he just asserts the point as though it were obvious.

For my money, this is an unresolvable debate.  Were it up to me, I would adopt the broader reading of  R.A.V., which Justice Scalia appears to reject. Under this broader reading, the First Amendment itself does not carve the world up into speech and non-speech. Instead, First Amendment scrutiny would apply whenever the government singles out acts because of hostility to the message they convey or takes action that substantially limits the ability of speakers to express their messages.  The second half of that test requires some judgment about what sorts of acts are usually engaged in for expressive purposes, but it could be folded into familiar doctrinal tests like whether a time, place or manner restriction leaves open adequate alternative channels of communication. I think that what I am proposing here is actually a more faithful reading of Justice Scalia's own opinion in R.A.V. than the position he and the majority of the Court stake out in Nevada Comm'n. It has the further virtue of largely avoiding the metaphysical question of what speech is.